See Full Example Code for full example code.
Part Mining from Online Repositories
In today's modern technological society, a variety of interesting technologies can be assembled from "off-the-shelf" components, including cars, computers, and airplanes. Synthetic biology is inspired by a similar idea. Synthetic biologists aim to program new biological functions into organisms by assembling genetic code from off-the-shelf DNA sequences. LibSBOL puts an inventory of biological parts at your fingertips.
For example, the iGEM Registry of Standard Biological Parts is an online resource that many synthetic biologists are familiar with. The Registry is an online database that catalogs a vast inventory of genetic parts, mostly contributed by students in the iGEM competition. These parts are now available in SBOL format in the SynBioHub knowledgebase, hosted by Newcastle University. The code example below demonstrates how a programmer can access these data.
The following code example shows how to pull data about biological components from the SynBioHub repository. The interface with the SynBioHub repository is represented by a PartShop object. The following code retrieves parts corresponding to promoter, coding sequence (CDS), ribosome binding site (RBS), and transcriptional terminator. These parts are imported into a Document object, which must be initialized first. See Getting Started with SBOL for more about creating Documents.
In order to pull a part, simply locate the web address of that part by browsing the SynBioHub repository online. Alternatively, libSBOL also supports programmatic querying of SynBioHub. LibSBOL supports three kinds of searches: a general search, an exact search, and an advanced search. The following query conducts a general search which scans through identity, name, description, and displayId properties for a match to the search text, including partial, case-insensitive matches to substrings of the property value. Search results are returned as a SearchResponse object.
By default, the general search looks only for ComponentDefinitions, and only returns 25 records at a time in order to prevent server overload. The search above is equivalent to the one below:
This request explicitly specifies which kind of SBOL object to search for, an offset of 0 (explained below), and a limit of 25 records. Of course, these default parameters can be changed to search for different type of SBOL objects or to return more records.
Some searches may match a large number of objects, more than the specified limit allows. The total number of objects matching the search criteria can be found using the searchCount method, which has the same call signature as the search method. In this case, it is possible to specify an offset and to retrieve additional records in successive requests. It is a good idea to put a small delay in between successive requests to prevent server overload. The following example demonstrates how to do this. The 100 millisecond delay is implemented using cross-platform C++11 headers chrono and thread. As of the writing of this documentation, this call retrieves 391 records.
Each record in a SearchResponse contains basic data, including identity, displayId, name, and description fields. It is very important to realize however that the search does not retrieve the complete ComponentDefinition! In order to retrieve the full object, the user must call pullComponentDefinition while specifying the target object's identity.
Records in a SearchResponse can be accessed using iterators or numeric indices. The interface for each record behaves exactly like any other SBOL object:
The preceding examples concern general searches, which scan through an object's metadata for partial matches to the search term. In contrast, the exact search explicitly specifies which property of an object to search, and the value of that property must exactly match the search term. The following exact search will search for ComponentDefinitions with a role of promoter:
Finally, the advanced search allows the user to configure a search with multiple criteria by constructing a SearchQuery object. The following query looks for promoters that have an additional annotation indicating that the promoter is regulated (as opposed to constitutive):
Computer-aided Design with SBOL
An advantage of the SBOL data format over GenBank is the ability to represent DNA as abstract components without specifying an exact sequence. An abstract design can be used as a template, with sequence information filled in later. In SBOL, a ComponentDefinition represents a biological component whose general function is known while its sequence is currently either unknown or unspecified. The intended function of the component is specified using a descriptive term from the Sequence Ontology (SO), a standard vocabulary for describing genetic parts. As the following example shows, some common SO terms are built in to libSBOL as pre-defined constants (see constants.h). This code example defines the new component as a gene. Other terms may be found by browsing the Sequence Ontology online.
Design abstraction is an important engineering principle for synthetic biology. Abstraction enables the engineer to think at a high-level about functional characteristics of a system while hiding low-level physical details. For example, in electronics, abstract schematics are used to describe the function of a circuit, while hiding the physical details of how a printed circuit board is laid out. Computer-aided design (CAD) programs allow the engineer to easily switch back and forth between abstract and physical representations of a circuit. In the same spirit, libSBOL enables a CAD approach for designing genetic constructs and other forms of synthetic biology.
Hierarchical DNA Assembly
LibSBOL also includes methods for assembling biological components into abstraction hierarchies. This is important from a biological perspective, because DNA sequences and biological structures in general exhibit hierarchical organization, from the genome, to operons, to genes, to lower level genetic operators. The following code assembles an abstraction hierarchy that describes a gene cassete. Note that subcomponents must belong to a Document in order to be assembled, so a Document is passed as a parameter.
After creating an abstraction hierarchy, it is then possible to iterate through an object's primary structure of components:
Caution! It is also possible to iterate through components as follows, but this way is not guaranteed to return components in sequential order. This is because SBOL supports a variety of structural descriptions, not just primary structure.
Sequence Assembly
A complete design adds explicit sequence information to the components in a template design. In order to complete a design, Sequence objects must first be created and associated with the promoter, CDS, RBS, terminator subcomponents. In contrast to the ComponentDefinition::assemble method, which assembles a template design, the Sequence::compile method recursively generates the complete sequence of a hierarchical design from the sequence of its subcomponents. Compiling a DNA sequence is analogous to a programmer compiling their code. In order to compile a Sequence, you must first assemble a template design from ComponentDefinitions, as described in the previous section.
Sequence Annotations
SequenceAnnotations describe regions of interest on Sequence objects. SequenceAnnotations are primarily used to describe sequence features that are not structural components of a design. Examples of SequenceAnnotations include start and stop codons, mutations, or restriction enzyme cut sites.
